Merge pull request #1623 from hkad98/jkd/metric-value-filter

feat(gooddata-sdk): add dimensionality support to MetricValueFilter

Author: hkad98

packages/gooddata-sdk/src/gooddata_sdk/compute/compute_to_sdk_converter.py

@@ -31,6 +31,14 @@
 from gooddata_sdk.utils import ref_extract, ref_extract_obj_id
 
 
+def _extract_dimensionality(f: dict[str, Any]) -> list[Union[str, ObjId, Attribute, Metric]] | None:
+    # mypy is unable to automatically convert Union[str, ObjId] to Union[str, ObjId, Attribute, Metric]
+    # so use explicit cast here
+    if "dimensionality" not in f:
+        return None
+    return [cast(Union[str, ObjId, Attribute, Metric], ref_extract(a)) for a in f["dimensionality"]]
+
+
 class ComputeToSdkConverter:
     """
     Provides functions to convert Compute API model objects represented as dictionaries to the SDK Compute model.
@@ -142,6 +150,7 @@ def convert_filter(filter_dict: dict[str, Any]) -> Filter:
                 operator=f["operator"],
                 values=f["value"],
                 treat_nulls_as=f.get("treatNullValuesAs"),
+                dimensionality=_extract_dimensionality(f),
             )
 
         if "rangeMeasureValueFilter" in filter_dict:
@@ -152,6 +161,7 @@ def convert_filter(filter_dict: dict[str, Any]) -> Filter:
                 operator=f["operator"],
                 values=(f["from"], f["to"]),
                 treat_nulls_as=f.get("treatNullValuesAs"),
+                dimensionality=_extract_dimensionality(f),
             )
 
         if "compoundMeasureValueFilter" in filter_dict:
@@ -174,22 +184,15 @@ def convert_filter(filter_dict: dict[str, Any]) -> Filter:
                 metric=ref_extract(f["measure"]),
                 conditions=conditions,
                 treat_nulls_as=f.get("treatNullValuesAs"),
+                dimensionality=_extract_dimensionality(f),
             )
 
         if "rankingFilter" in filter_dict:
             f = filter_dict["rankingFilter"]
 
-            # mypy is unable to automatically convert Union[str, ObjId] to Union[str, ObjId, Attribute, Metric]
-            # so use explicit cast here
-            dimensionality = (
-                [cast(Union[str, ObjId, Attribute, Metric], ref_extract(a)) for a in f["dimensionality"]]
-                if "dimensionality" in f
-                else None
-            )
-
             return RankingFilter(
                 metrics=[ref_extract(m) for m in f["measures"]],
-                dimensionality=dimensionality,
+                dimensionality=_extract_dimensionality(f),
                 operator=f["operator"],
                 value=f["value"],
             )

packages/gooddata-sdk/src/gooddata_sdk/compute/model/filter.py

@@ -586,6 +586,7 @@ def __init__(
         operator: str,
         values: Union[float, int, tuple[float, float]],
         treat_nulls_as: Union[float, None] = None,
+        dimensionality: list[Union[str, ObjId, Attribute, Metric]] | None = None,
     ) -> None:
         super().__init__()
 
@@ -611,6 +612,7 @@ def __init__(
         self._metric = _extract_id_or_local_id(metric)
         self._operator = operator
         self._treat_nulls_as = treat_nulls_as
+        self._dimensionality = [_extract_id_or_local_id(d) for d in dimensionality] if dimensionality else None
 
     @property
     def metric(self) -> Union[ObjId, str]:
@@ -632,19 +634,25 @@ def values(self) -> Union[tuple[float], tuple[float, float]]:
     def treat_nulls_as(self) -> Union[float, None]:
         return self._treat_nulls_as
 
+    @property
+    def dimensionality(self) -> list[Union[ObjId, str]] | None:
+        return self._dimensionality
+
     def is_noop(self) -> bool:
         return False
 
     def as_api_model(self) -> Union[afm_models.ComparisonMeasureValueFilter, afm_models.RangeMeasureValueFilter]:
         measure = _to_identifier(self._metric)
 
-        kwargs = dict(
+        kwargs: dict[str, Any] = dict(
             measure=measure,
             operator=self.operator,
             _check_type=False,
         )
         if self.treat_nulls_as is not None:
             kwargs["treat_null_values_as"] = self.treat_nulls_as
+        if self._dimensionality is not None:
+            kwargs["dimensionality"] = [_to_identifier(d) for d in self._dimensionality]
 
         if _METRIC_VALUE_FILTER_OPERATORS[self.operator] == "comparison":
             kwargs["value"] = self.values[0]
@@ -727,11 +735,13 @@ def __init__(
         metric: Union[ObjId, str, Metric],
         conditions: list[MetricValueCondition],
         treat_nulls_as: Union[float, None] = None,
+        dimensionality: list[Union[str, ObjId, Attribute, Metric]] | None = None,
     ) -> None:
         super().__init__()
         self._metric = _extract_id_or_local_id(metric)
         self._conditions = conditions
         self._treat_nulls_as = treat_nulls_as
+        self._dimensionality = [_extract_id_or_local_id(d) for d in dimensionality] if dimensionality else None
 
     @property
     def metric(self) -> Union[ObjId, str]:
@@ -745,6 +755,10 @@ def conditions(self) -> list[MetricValueCondition]:
     def treat_nulls_as(self) -> Union[float, None]:
         return self._treat_nulls_as
 
+    @property
+    def dimensionality(self) -> list[Union[ObjId, str]] | None:
+        return self._dimensionality
+
     def is_noop(self) -> bool:
         return len(self.conditions) == 0
 
@@ -758,6 +772,8 @@ def as_api_model(self) -> afm_models.CompoundMeasureValueFilter:
         )
         if self.treat_nulls_as is not None:
             kwargs["treat_null_values_as"] = self.treat_nulls_as
+        if self._dimensionality is not None:
+            kwargs["dimensionality"] = [_to_identifier(d) for d in self._dimensionality]
 
         body = CompoundMeasureValueFilterBody(**kwargs)
         return afm_models.CompoundMeasureValueFilter(body, _check_type=False)

packages/gooddata-sdk/tests/compute/test_compute_to_sdk_converter.py

@@ -13,6 +13,7 @@
     MetricValueFilter,
     MetricValueRangeCondition,
     NegativeAttributeFilter,
+    ObjId,
     PopDateMetric,
     PopDatesetMetric,
     PositiveAttributeFilter,
@@ -270,6 +271,77 @@ def test_compound_measure_value_filter_conversion():
     assert result.conditions[1].to_value == 20
 
 
+def test_comparison_measure_value_filter_with_dimensionality_conversion():
+    filter_dict = json.loads(
+        """
+        {
+          "comparisonMeasureValueFilter": {
+            "measure": { "localIdentifier": "measureLocalId" },
+            "operator": "GREATER_THAN",
+            "value": 100,
+            "dimensionality": [
+                { "localIdentifier": "attributeLocalId" },
+                { "identifier": { "id": "label.id", "type": "label" } }
+            ]
+          }
+        }
+        """
+    )
+
+    result = ComputeToSdkConverter.convert_filter(filter_dict)
+
+    assert isinstance(result, MetricValueFilter)
+    assert result.dimensionality is not None
+    assert result.dimensionality[0] == "attributeLocalId"
+    assert result.dimensionality[1] == ObjId(type="label", id="label.id")
+
+
+def test_range_measure_value_filter_with_dimensionality_conversion():
+    filter_dict = json.loads(
+        """
+        {
+          "rangeMeasureValueFilter": {
+            "measure": { "localIdentifier": "measureLocalId" },
+            "operator": "BETWEEN",
+            "from": 100,
+            "to": 200,
+            "dimensionality": [
+                { "localIdentifier": "attributeLocalId" }
+            ]
+          }
+        }
+        """
+    )
+
+    result = ComputeToSdkConverter.convert_filter(filter_dict)
+
+    assert isinstance(result, MetricValueFilter)
+    assert result.dimensionality == ["attributeLocalId"]
+
+
+def test_compound_measure_value_filter_with_dimensionality_conversion():
+    filter_dict = json.loads(
+        """
+        {
+          "compoundMeasureValueFilter": {
+            "measure": { "localIdentifier": "measureLocalId" },
+            "conditions": [
+              { "comparison": { "operator": "GREATER_THAN", "value": 100 } }
+            ],
+            "dimensionality": [
+                { "localIdentifier": "attributeLocalId" }
+            ]
+          }
+        }
+        """
+    )
+
+    result = ComputeToSdkConverter.convert_filter(filter_dict)
+
+    assert isinstance(result, CompoundMetricValueFilter)
+    assert result.dimensionality == ["attributeLocalId"]
+
+
 def test_ranking_filter_conversion():
     filter_dict = json.loads(
         """

packages/gooddata-sdk/tests/compute_model/metric_value_filter/comparison_filter_with_dimensionality_and_treat_nulls_as.snapshot.json

@@ -0,0 +1,15 @@
+{
+    "comparison_measure_value_filter": {
+        "dimensionality": [
+            {
+                "local_identifier": "local_id3"
+            }
+        ],
+        "measure": {
+            "local_identifier": "local_id1"
+        },
+        "operator": "GREATER_THAN",
+        "treat_null_values_as": 0,
+        "value": 10.0
+    }
+}

packages/gooddata-sdk/tests/compute_model/metric_value_filter/comparison_filter_with_dimensionality_using_local_ids.snapshot.json

@@ -0,0 +1,17 @@
+{
+    "comparison_measure_value_filter": {
+        "dimensionality": [
+            {
+                "local_identifier": "local_id3"
+            },
+            {
+                "local_identifier": "local_id4"
+            }
+        ],
+        "measure": {
+            "local_identifier": "local_id1"
+        },
+        "operator": "GREATER_THAN",
+        "value": 10.0
+    }
+}

packages/gooddata-sdk/tests/compute_model/metric_value_filter/comparison_filter_with_dimensionality_using_mix_of_ids.snapshot.json

@@ -0,0 +1,20 @@
+{
+    "comparison_measure_value_filter": {
+        "dimensionality": [
+            {
+                "local_identifier": "local_id3"
+            },
+            {
+                "identifier": {
+                    "id": "label.id",
+                    "type": "label"
+                }
+            }
+        ],
+        "measure": {
+            "local_identifier": "local_id1"
+        },
+        "operator": "GREATER_THAN",
+        "value": 10.0
+    }
+}

packages/gooddata-sdk/tests/compute_model/metric_value_filter/range_filter_with_dimensionality.snapshot.json

@@ -0,0 +1,18 @@
+{
+    "range_measure_value_filter": {
+        "_from": 2,
+        "dimensionality": [
+            {
+                "local_identifier": "local_id3"
+            },
+            {
+                "local_identifier": "local_id4"
+            }
+        ],
+        "measure": {
+            "local_identifier": "local_id1"
+        },
+        "operator": "BETWEEN",
+        "to": 3
+    }
+}

packages/gooddata-sdk/tests/compute_model/test_compound_metric_value_filter.py

@@ -2,6 +2,7 @@
 from __future__ import annotations
 
 from gooddata_sdk import (
+    Attribute,
     CompoundMetricValueFilter,
     MetricValueComparisonCondition,
     MetricValueRangeCondition,
@@ -35,3 +36,64 @@ def test_compound_metric_value_filter_to_api_model():
 def test_compound_metric_value_filter_noop_when_no_conditions():
     f = CompoundMetricValueFilter(metric=ObjId(type="metric", id="metric.id"), conditions=[])
     assert f.is_noop() is True
+
+
+def test_compound_metric_value_filter_with_dimensionality_to_api_model():
+    f = CompoundMetricValueFilter(
+        metric="local_id1",
+        conditions=[
+            MetricValueComparisonCondition(operator="GREATER_THAN", value=10),
+        ],
+        dimensionality=["local_id3", Attribute(local_id="local_id4", label="label.id")],
+    )
+
+    assert f.dimensionality == ["local_id3", "local_id4"]
+    assert f.as_api_model().to_dict() == {
+        "compound_measure_value_filter": {
+            "conditions": [
+                {"comparison": {"operator": "GREATER_THAN", "value": 10.0}},
+            ],
+            "dimensionality": [
+                {"local_identifier": "local_id3"},
+                {"local_identifier": "local_id4"},
+            ],
+            "measure": {"local_identifier": "local_id1"},
+        }
+    }
+
+
+def test_compound_metric_value_filter_with_dimensionality_mixed_ids():
+    f = CompoundMetricValueFilter(
+        metric="local_id1",
+        conditions=[
+            MetricValueComparisonCondition(operator="GREATER_THAN", value=10),
+        ],
+        dimensionality=["local_id3", ObjId(type="label", id="label.id")],
+        treat_nulls_as=0,
+    )
+
+    assert f.as_api_model().to_dict() == {
+        "compound_measure_value_filter": {
+            "conditions": [
+                {"comparison": {"operator": "GREATER_THAN", "value": 10.0}},
+            ],
+            "dimensionality": [
+                {"local_identifier": "local_id3"},
+                {"identifier": {"id": "label.id", "type": "label"}},
+            ],
+            "measure": {"local_identifier": "local_id1"},
+            "treat_null_values_as": 0,
+        }
+    }
+
+
+def test_compound_metric_value_filter_without_dimensionality_omits_field():
+    f = CompoundMetricValueFilter(
+        metric="local_id1",
+        conditions=[
+            MetricValueComparisonCondition(operator="GREATER_THAN", value=10),
+        ],
+    )
+
+    assert f.dimensionality is None
+    assert "dimensionality" not in f.as_api_model().to_dict()["compound_measure_value_filter"]